Impacts of model resolution on predictions of air quality and associated health exposure in Nanjing, China

Air quality models have been used in health studies to provide spatial and temporal information of various air pollutants. Model resolution is an important factor affecting the accuracy of exposure assessment using model predictions. In this study, the WRF/CMAQ model system was applied to quantitatively estimate the impacts of the model resolution on the predictions of air quality and associated health exposure in Nanjing, China in 2016. Air quality was simulated with a grid resolution of 1, 4, 12, and 36 km respectively. Predictions with 1 or 4 km resolution are slightly better for particulate matter with an aerodynamic diameter <= 2.5 mu m (PM2.5) and its compositions and predictions with 12 km are slightly better for daily 8-h maximum ozone (O-3-8 h). Model resolution does not significantly improve predictions for PM2.5 and O-3-8 h in Nanjing, however, the spatial distributions of PM2.5 and O-3-8 h are better captured with finer resolutions. Population weighted concentrations (PWCs) of PM2.5 with different model resolutions are similar to the average of observations, but PWCs of O-3-8 h with all resolutions are obviously larger than the observations, indicating that the current sites may well represent the population exposure to PM2.5, but under-estimate the exposure to O-3. Model resolution results in about 6% in the estimated premature mortality due to exposure to PM2.5 but more than 20% difference in premature mortality due to exposure to O-3. Future studies are needed to evaluate the impacts of the resolution on the exposure of PM2.5 compositions in the city scale when PM2.5 composition measurements available at multiple sites. (C) 2020 Elsevier Ltd. All rights reserved.